1
00:00:04,120 --> 00:00:07,160
Speaker 1: Get in touch with technology with tech Stuff from how

2
00:00:07,200 --> 00:00:13,880
Speaker 1: stuff Works dot com. Hey there, and welcome to tech Stuff.

3
00:00:13,880 --> 00:00:17,959
Speaker 1: I'm your host, Jonathan Strickland. I'm an executive producer and

4
00:00:18,079 --> 00:00:21,560
Speaker 1: I love all things tech. And today we're gonna talk

5
00:00:21,600 --> 00:00:28,440
Speaker 1: about the concept of granting personhood or personality two robots,

6
00:00:28,480 --> 00:00:31,280
Speaker 1: not a personality in the sense of this robot is

7
00:00:31,360 --> 00:00:35,640
Speaker 1: Chipper and this one's depressed, but rather the concept of

8
00:00:35,880 --> 00:00:42,800
Speaker 1: granting robots some of the aspects that we say humans

9
00:00:43,000 --> 00:00:48,720
Speaker 1: have from a legal standpoint, and there are numerous arguments

10
00:00:48,760 --> 00:00:54,800
Speaker 1: both for and against this concept. Uh. I've talked about

11
00:00:54,960 --> 00:00:59,000
Speaker 1: artificial intelligence extensively, not just on this show but on

12
00:00:59,120 --> 00:01:03,040
Speaker 1: other shows as well. And one of the elements about

13
00:01:03,120 --> 00:01:07,319
Speaker 1: artificial intelligence that tends to pop up, especially in things

14
00:01:07,319 --> 00:01:12,440
Speaker 1: like science fiction, is what happens when artificial intelligence is

15
00:01:12,480 --> 00:01:19,040
Speaker 1: able to take actions that could negatively impact people? Uh,

16
00:01:19,120 --> 00:01:21,520
Speaker 1: what do we do? What sort of framework do we

17
00:01:21,600 --> 00:01:26,160
Speaker 1: make if that were the case, where we can determine

18
00:01:26,240 --> 00:01:29,319
Speaker 1: who's responsible for this? And then the interesting thing is

19
00:01:29,360 --> 00:01:34,759
Speaker 1: that for me anyway, the European Parliament, a group out

20
00:01:34,760 --> 00:01:37,440
Speaker 1: of the European Parliament kind of a working group, put

21
00:01:37,480 --> 00:01:41,800
Speaker 1: together a proposal a couple of years ago about the

22
00:01:41,840 --> 00:01:45,240
Speaker 1: possibility of this and the idea of granting personhood to

23
00:01:45,440 --> 00:01:49,400
Speaker 1: robots in a way to create an established framework of

24
00:01:49,480 --> 00:01:54,760
Speaker 1: law and accountability, because at the moment there's not really

25
00:01:54,800 --> 00:01:58,800
Speaker 1: anything formalized there, so they put together a proposal. The

26
00:01:58,840 --> 00:02:04,520
Speaker 1: proposal UH is still online. It was originally published on

27
00:02:04,720 --> 00:02:08,000
Speaker 1: May thirty one, two thousand sixteen. It was submitted in

28
00:02:08,080 --> 00:02:11,720
Speaker 1: draft form. The title of the draft was Draft Report

29
00:02:11,800 --> 00:02:15,840
Speaker 1: with Recommendations to the Commission on Civil Law Rules on

30
00:02:16,080 --> 00:02:20,560
Speaker 1: Robotics and it's a fascinating report. It's actually a lot

31
00:02:20,560 --> 00:02:22,799
Speaker 1: of fun to read. I highly recommend checking it out

32
00:02:22,880 --> 00:02:26,240
Speaker 1: if you have some time. The general purpose of the

33
00:02:26,280 --> 00:02:31,280
Speaker 1: proposal was to start official discussions in the European Union

34
00:02:31,760 --> 00:02:36,040
Speaker 1: about developing policies and guidelines in the field of robotics,

35
00:02:36,080 --> 00:02:41,639
Speaker 1: particularly since there were individual member states of the EU

36
00:02:41,760 --> 00:02:45,799
Speaker 1: s several nations that were developing their own policies over time,

37
00:02:46,639 --> 00:02:51,080
Speaker 1: and the European Parliament Working Group was saying this could

38
00:02:51,080 --> 00:02:54,920
Speaker 1: be a problem because if we have one set of

39
00:02:54,960 --> 00:02:59,440
Speaker 1: policies that are in play and let's say France, and

40
00:02:59,480 --> 00:03:03,040
Speaker 1: a different set that are in play in Germany, that

41
00:03:03,560 --> 00:03:07,600
Speaker 1: starts to create conflict, and in the European Union, you're

42
00:03:07,600 --> 00:03:10,280
Speaker 1: supposed to be able to move throughout the union and

43
00:03:10,320 --> 00:03:14,799
Speaker 1: take jobs wherever and live wherever within the Union. And

44
00:03:15,360 --> 00:03:19,120
Speaker 1: therefore it would work a lot better if everyone had

45
00:03:19,160 --> 00:03:22,960
Speaker 1: the same set of rules and regulations on something like this,

46
00:03:23,080 --> 00:03:28,200
Speaker 1: especially something that's going to factor so heavily in industry

47
00:03:28,360 --> 00:03:33,480
Speaker 1: and economy in the workplace. So really this was mostly

48
00:03:33,520 --> 00:03:36,560
Speaker 1: a kind of a warning saying we need to start

49
00:03:36,560 --> 00:03:40,760
Speaker 1: talking about this. The introduction of that proposal is phenomenal

50
00:03:40,800 --> 00:03:43,960
Speaker 1: in its own right. It cites lots of literary works,

51
00:03:44,000 --> 00:03:50,480
Speaker 1: including Mary Shelley for Frankenstein, The Pygmalion, myth Uh. It

52
00:03:51,120 --> 00:03:56,920
Speaker 1: references Carol Capex Are You Are? That is the play

53
00:03:56,960 --> 00:04:03,520
Speaker 1: that originated the word robot. So it's pretty interesting just

54
00:04:03,680 --> 00:04:07,760
Speaker 1: from the fact that it's referencing science fiction and horror

55
00:04:08,320 --> 00:04:12,320
Speaker 1: literature way more than you would expect in your typical

56
00:04:13,000 --> 00:04:17,480
Speaker 1: government proposal. It's it's much more entertaining than reading a

57
00:04:18,279 --> 00:04:22,560
Speaker 1: draft of, say, uh like your typical agricultural report, which

58
00:04:22,560 --> 00:04:25,520
Speaker 1: is still incredibly important. Don't get me wrong, those are

59
00:04:25,720 --> 00:04:29,960
Speaker 1: very important reports. They just don't tend to be page turners. Now,

60
00:04:29,960 --> 00:04:33,240
Speaker 1: those citations are all there to establish that we humans

61
00:04:33,279 --> 00:04:38,840
Speaker 1: have been really fascinated, maybe even fixated with this idea

62
00:04:39,400 --> 00:04:43,560
Speaker 1: of creating intelligent machines or even intelligent life in the

63
00:04:43,600 --> 00:04:47,520
Speaker 1: case of Frankenstein's Monster, that this has been something that

64
00:04:47,600 --> 00:04:52,599
Speaker 1: we have really aspired to on some level, the idea

65
00:04:52,640 --> 00:04:56,320
Speaker 1: that we create something that itself could be said to

66
00:04:56,440 --> 00:05:00,960
Speaker 1: be intelligent. Section B of the introduction, as the introductions

67
00:05:01,000 --> 00:05:04,400
Speaker 1: long enough to have different sections, would say that we're

68
00:05:04,440 --> 00:05:09,279
Speaker 1: on the threshold of a new industrial revolution. That the

69
00:05:09,320 --> 00:05:13,480
Speaker 1: first industrial revolution that took place in the nineteenth century

70
00:05:13,800 --> 00:05:20,360
Speaker 1: was all about creating factories, harnessing the power of coal building, railroads,

71
00:05:20,960 --> 00:05:25,960
Speaker 1: steam power, assembly lines, all of that stuff transformed the

72
00:05:26,000 --> 00:05:28,760
Speaker 1: world from what what had been a really a grarian

73
00:05:28,960 --> 00:05:33,240
Speaker 1: society into more of an urban one in these industrialized nations.

74
00:05:34,240 --> 00:05:38,960
Speaker 1: And the section was saying that artificial intelligence was going

75
00:05:39,000 --> 00:05:42,520
Speaker 1: to fuel a new revolution that would be just as

76
00:05:42,600 --> 00:05:46,880
Speaker 1: impactful as the previous industrial revolution, that no part of

77
00:05:46,960 --> 00:05:51,239
Speaker 1: society would go unaffected by that revolution. So therefore, since

78
00:05:51,240 --> 00:05:54,599
Speaker 1: it was going to be such a big part of

79
00:05:54,600 --> 00:05:57,440
Speaker 1: our lives moving forward, it would be really smart for

80
00:05:57,520 --> 00:06:00,839
Speaker 1: us to think about the implications of that out before

81
00:06:00,880 --> 00:06:04,359
Speaker 1: it happens, and prepare for it and to put into

82
00:06:04,440 --> 00:06:09,520
Speaker 1: place protections for people before it becomes reactionary, you know,

83
00:06:09,600 --> 00:06:11,960
Speaker 1: before there's a problem and then we have to figure out, oh,

84
00:06:12,000 --> 00:06:15,359
Speaker 1: how can we fix this. The argument of the proposal was,

85
00:06:15,560 --> 00:06:19,440
Speaker 1: let's be proactive, let's try and figure out what problems

86
00:06:19,520 --> 00:06:24,200
Speaker 1: may confront us in the future and solve them now

87
00:06:24,800 --> 00:06:31,039
Speaker 1: before we are dealing with a catastrophic experience. So then

88
00:06:31,080 --> 00:06:34,240
Speaker 1: the proposal goes and points out that robots sales have

89
00:06:34,320 --> 00:06:36,640
Speaker 1: been on the rise over the past few years. They

90
00:06:36,760 --> 00:06:40,480
Speaker 1: they've increased year over year. The automotive industry in particular

91
00:06:40,880 --> 00:06:44,080
Speaker 1: was called out, and it also stated that robots will

92
00:06:44,120 --> 00:06:48,120
Speaker 1: provide numerous benefits in the short to medium term and

93
00:06:48,160 --> 00:06:54,040
Speaker 1: potentially what they called virtually unbounded prosperity in the long term.

94
00:06:54,120 --> 00:06:58,080
Speaker 1: So the idea being there are some real positive outcomes

95
00:06:58,200 --> 00:07:02,640
Speaker 1: to employing robots. However, the flip side of that is

96
00:07:02,680 --> 00:07:06,600
Speaker 1: those advances might quote result in a large part of

97
00:07:06,600 --> 00:07:09,240
Speaker 1: the work now done by humans being taken over by

98
00:07:09,320 --> 00:07:12,760
Speaker 1: robots end quote, and that that would affect not just employment,

99
00:07:12,840 --> 00:07:16,720
Speaker 1: but systems like social security which rely upon employment taxes

100
00:07:16,720 --> 00:07:20,600
Speaker 1: for funding and other things that are tax supported. And

101
00:07:20,680 --> 00:07:22,840
Speaker 1: what happens when a robot fails in some way? What

102
00:07:22,880 --> 00:07:26,360
Speaker 1: if it creates damage to a person or property? And

103
00:07:26,640 --> 00:07:30,320
Speaker 1: how will robots interact with our personal data, even you know,

104
00:07:30,600 --> 00:07:34,400
Speaker 1: personal data that we haven't necessarily given consent to share.

105
00:07:35,160 --> 00:07:38,200
Speaker 1: The more intelligent the system, the more proactive it might

106
00:07:38,240 --> 00:07:41,640
Speaker 1: be in taking that data and using it in some

107
00:07:41,720 --> 00:07:45,120
Speaker 1: way without us even saying anything about it, especially if

108
00:07:45,160 --> 00:07:47,680
Speaker 1: you haven't built in specific rules for the robot or

109
00:07:47,760 --> 00:07:51,240
Speaker 1: artificial intelligent program or whatever to follow so that it

110
00:07:51,320 --> 00:07:55,000
Speaker 1: doesn't do that. Now, before I go much further down

111
00:07:55,080 --> 00:07:58,160
Speaker 1: that particular line of reasoning, I do need to point

112
00:07:58,160 --> 00:08:03,080
Speaker 1: out that AI and automation and their effect on jobs

113
00:08:03,360 --> 00:08:07,520
Speaker 1: is still a matter of considerable debate. No one is

114
00:08:07,560 --> 00:08:10,480
Speaker 1: sure right now to what extent it's going to have

115
00:08:10,600 --> 00:08:14,160
Speaker 1: an effect on people in general. So the worst case

116
00:08:14,160 --> 00:08:18,160
Speaker 1: scenario that some people say is that robots and AI

117
00:08:18,280 --> 00:08:21,760
Speaker 1: automated systems are going to replace the vast majority of

118
00:08:21,880 --> 00:08:25,960
Speaker 1: jobs within a few decades, and that this will happen

119
00:08:26,080 --> 00:08:29,440
Speaker 1: before we've ever built out any sort of infrastructure that

120
00:08:29,480 --> 00:08:33,040
Speaker 1: would take care of people that would help them transition

121
00:08:33,440 --> 00:08:37,280
Speaker 1: either into new jobs that were created as a result

122
00:08:37,400 --> 00:08:40,839
Speaker 1: of this, jobs that we don't have today because they

123
00:08:41,320 --> 00:08:44,000
Speaker 1: we don't need them, but we may have a need

124
00:08:44,040 --> 00:08:48,480
Speaker 1: of them once this automated future comes upon us. Or

125
00:08:49,600 --> 00:08:52,760
Speaker 1: maybe that isn't the case. Maybe we don't create new jobs,

126
00:08:53,080 --> 00:08:56,800
Speaker 1: Maybe we don't have a system where we can actually

127
00:08:57,200 --> 00:09:01,800
Speaker 1: take care of people and and separate the concepts of

128
00:09:02,080 --> 00:09:06,960
Speaker 1: making a living and being employed, having some other system

129
00:09:07,040 --> 00:09:10,720
Speaker 1: there so that people could meet their needs to make

130
00:09:10,880 --> 00:09:14,080
Speaker 1: ends meet. How do we do that without having it

131
00:09:14,120 --> 00:09:18,440
Speaker 1: tied to jobs? But there are other people who also

132
00:09:18,480 --> 00:09:22,160
Speaker 1: think we're not likely to see that huge of a change,

133
00:09:22,200 --> 00:09:27,040
Speaker 1: at least not in the relatively near future. We're more

134
00:09:27,160 --> 00:09:31,240
Speaker 1: likely going to see AI and automation take over individual tasks,

135
00:09:31,720 --> 00:09:35,240
Speaker 1: but not entire jobs. So the future might be one

136
00:09:35,280 --> 00:09:38,360
Speaker 1: in which our work is augmented by AI, but we

137
00:09:38,440 --> 00:09:43,200
Speaker 1: are not replaced by AI. The justification for that argument

138
00:09:43,280 --> 00:09:46,520
Speaker 1: is saying, look at machine learning as it stands right now.

139
00:09:46,559 --> 00:09:50,080
Speaker 1: It is very impressive, but it also shows that there

140
00:09:50,080 --> 00:09:54,360
Speaker 1: are severe limitations in machine learning right now. It's just

141
00:09:54,440 --> 00:09:58,480
Speaker 1: not sophisticated enough to be able to take over for

142
00:09:58,600 --> 00:10:01,760
Speaker 1: everything that a human can do. Uh, and it requires

143
00:10:01,800 --> 00:10:03,440
Speaker 1: a lot of training and a lot of stuff can

144
00:10:03,480 --> 00:10:07,200
Speaker 1: go wrong, So it's not likely that we're gonna see

145
00:10:07,200 --> 00:10:10,240
Speaker 1: computers take over lots and lots and lots of jobs,

146
00:10:10,240 --> 00:10:13,280
Speaker 1: but they might take over repetitive tasks that your job

147
00:10:13,480 --> 00:10:16,319
Speaker 1: happens to include, and then you can focus on the

148
00:10:16,360 --> 00:10:21,120
Speaker 1: stuff that isn't repetitive and predictable. Who's to say, who's

149
00:10:21,200 --> 00:10:24,440
Speaker 1: right not me? I don't know. It could be either

150
00:10:24,480 --> 00:10:26,280
Speaker 1: one of those. But the argument of the proposal I

151
00:10:26,320 --> 00:10:28,840
Speaker 1: think still holds true, which is that we need to

152
00:10:29,559 --> 00:10:33,160
Speaker 1: at least consider what the worst case scenario is and

153
00:10:33,440 --> 00:10:38,880
Speaker 1: have a plan to alleviate the outcome of that. Then

154
00:10:38,920 --> 00:10:41,520
Speaker 1: there's the question of how robots are going to impact

155
00:10:41,720 --> 00:10:45,120
Speaker 1: stuff like human dignity. So in the future, if we

156
00:10:45,200 --> 00:10:48,559
Speaker 1: have robots that are acting like caretakers for the sick

157
00:10:48,960 --> 00:10:51,600
Speaker 1: or the elderly or the young, what impact is that

158
00:10:51,640 --> 00:10:54,920
Speaker 1: going to have on those people, How is that going

159
00:10:54,960 --> 00:10:57,880
Speaker 1: to make them feel? What impact will have on their health?

160
00:10:58,679 --> 00:11:01,200
Speaker 1: And uh, and hey, what if those robots were to

161
00:11:01,240 --> 00:11:07,320
Speaker 1: get really effective, really smart, like maybe smarter than humans smart,

162
00:11:07,840 --> 00:11:12,199
Speaker 1: So again, not necessarily intelligent in the same way humans are,

163
00:11:12,240 --> 00:11:15,199
Speaker 1: but better able to process certain types of information than

164
00:11:15,400 --> 00:11:20,720
Speaker 1: humans are. The proposal in the introduction, it raises that

165
00:11:20,800 --> 00:11:23,800
Speaker 1: question as well. Could robots actually represent a danger to

166
00:11:23,840 --> 00:11:26,840
Speaker 1: the human species. It's something that we have to consider

167
00:11:27,400 --> 00:11:31,360
Speaker 1: before it becomes a reality. Now, with all of this

168
00:11:31,480 --> 00:11:35,360
Speaker 1: in mind, says the introduction, the EU should get off

169
00:11:35,360 --> 00:11:38,880
Speaker 1: its butt and start talking about those ideas and work

170
00:11:38,880 --> 00:11:42,760
Speaker 1: on a strategy to avoid problems in the future. And

171
00:11:43,280 --> 00:11:46,320
Speaker 1: the proposal would then go on to have a few

172
00:11:46,360 --> 00:11:48,480
Speaker 1: suggestions of its own. I'll get into those in just

173
00:11:48,600 --> 00:11:51,760
Speaker 1: a second, but first let's take a quick break to

174
00:11:51,880 --> 00:12:03,480
Speaker 1: thank our sponsor. So the proposal has a solutions and

175
00:12:03,559 --> 00:12:08,760
Speaker 1: Suggestions section, and under general principles, the report would cite

176
00:12:08,840 --> 00:12:14,600
Speaker 1: Asimov's Laws of Robotics and would say that designers, producers,

177
00:12:14,600 --> 00:12:18,600
Speaker 1: and operators of robots should keep these in mind, which

178
00:12:18,679 --> 00:12:22,720
Speaker 1: is amusing to me. I mean asthma asthmava. Isaac Asimov

179
00:12:23,000 --> 00:12:29,120
Speaker 1: was a famed science fiction and speculative fiction author wrote

180
00:12:29,160 --> 00:12:33,160
Speaker 1: some amazing stuff, and those laws of robotics have become

181
00:12:33,200 --> 00:12:38,680
Speaker 1: sort of iconic in artificial intelligence. The original three laws

182
00:12:38,720 --> 00:12:41,959
Speaker 1: of robotics are probably the best known. So law number

183
00:12:42,000 --> 00:12:44,880
Speaker 1: one is a robot may not harm a human or

184
00:12:45,120 --> 00:12:48,599
Speaker 1: through in action, allow a human to come to harm.

185
00:12:48,840 --> 00:12:52,439
Speaker 1: Law too is a robot has to obey any order

186
00:12:52,800 --> 00:12:56,480
Speaker 1: given to it by a human unless it would violate

187
00:12:56,720 --> 00:13:00,000
Speaker 1: the first law. So you could tell a robot, hey

188
00:13:00,000 --> 00:13:01,440
Speaker 1: will pick that up for me, and it would have

189
00:13:01,520 --> 00:13:04,400
Speaker 1: to do it. But you couldn't tell robot, hey, go

190
00:13:04,600 --> 00:13:07,360
Speaker 1: push that guy into traffic, because that would violate the

191
00:13:07,400 --> 00:13:10,600
Speaker 1: first law. The third law was a robot would have

192
00:13:10,679 --> 00:13:15,280
Speaker 1: to protect itself from harm unless doing so would conflict

193
00:13:15,320 --> 00:13:19,400
Speaker 1: with either of the first two laws. So if a

194
00:13:19,559 --> 00:13:23,600
Speaker 1: robot were to see that a car were coming, it

195
00:13:23,640 --> 00:13:25,439
Speaker 1: was coming down the street, it was gonna hit a

196
00:13:25,480 --> 00:13:27,920
Speaker 1: little old lady, and the robot would be able to

197
00:13:27,960 --> 00:13:30,559
Speaker 1: push the little lady out of the way, but as

198
00:13:30,559 --> 00:13:33,120
Speaker 1: a result, it was going to get hit by this car.

199
00:13:33,320 --> 00:13:35,400
Speaker 1: The robot would have to do it because even though

200
00:13:35,440 --> 00:13:38,240
Speaker 1: it has a law stating it has to protect itself,

201
00:13:38,760 --> 00:13:41,960
Speaker 1: it that gets superseded by law number one, which says

202
00:13:42,480 --> 00:13:44,920
Speaker 1: it cannot through in action, allow a human to come

203
00:13:44,960 --> 00:13:47,160
Speaker 1: to harm. It would have to take action in that case.

204
00:13:47,880 --> 00:13:51,520
Speaker 1: Then there's a fourth law called law zero that was

205
00:13:51,559 --> 00:13:55,360
Speaker 1: considered to be the top of the this ladder, the

206
00:13:55,480 --> 00:13:57,920
Speaker 1: most important of all the laws, which is a robot

207
00:13:57,960 --> 00:14:02,280
Speaker 1: may not harm humanity or by inaction allow humanity to

208
00:14:02,320 --> 00:14:04,960
Speaker 1: come to harm, so not just an individual, but humanity

209
00:14:05,080 --> 00:14:09,880
Speaker 1: as a whole. Under the liability section of the report,

210
00:14:10,280 --> 00:14:12,960
Speaker 1: it suggests that it will not be long before the

211
00:14:13,000 --> 00:14:17,000
Speaker 1: European Union either needs to classify robots under a category

212
00:14:17,040 --> 00:14:20,400
Speaker 1: such as persons or to create a brand new category

213
00:14:20,520 --> 00:14:25,240
Speaker 1: just for robots. So AI and machine learning are really

214
00:14:25,640 --> 00:14:31,880
Speaker 1: changing how robots interact with environments. Before you would program

215
00:14:31,920 --> 00:14:35,960
Speaker 1: all the ways that a robot would interact with its environment,

216
00:14:36,080 --> 00:14:40,160
Speaker 1: and mostly you would try to control the types of

217
00:14:40,240 --> 00:14:42,680
Speaker 1: environments your robot was going to be in. You know,

218
00:14:42,760 --> 00:14:46,120
Speaker 1: the more predictable and the more stationary the robot, the

219
00:14:46,160 --> 00:14:48,240
Speaker 1: easier it was to program. Right, So if you've got

220
00:14:48,240 --> 00:14:51,280
Speaker 1: a robot, like a giant robotic arm that's doing welding

221
00:14:51,400 --> 00:14:54,640
Speaker 1: in a car manufacturing line, that robot is going to

222
00:14:54,680 --> 00:14:58,080
Speaker 1: be stationary. It doesn't move around. It stays in one

223
00:14:58,120 --> 00:15:00,840
Speaker 1: spot along the assembly line. The car has come to it,

224
00:15:00,840 --> 00:15:05,200
Speaker 1: it does its work, next car comes to it, it continues. Uh.

225
00:15:05,280 --> 00:15:09,600
Speaker 1: That is one way of programming a robot, and it's

226
00:15:09,800 --> 00:15:14,440
Speaker 1: easy comparatively speaking. But these days we now have robots

227
00:15:14,480 --> 00:15:19,960
Speaker 1: that use machine learning that encounter situations and then process

228
00:15:20,000 --> 00:15:24,320
Speaker 1: information come up with a conclusion. It might be a

229
00:15:24,320 --> 00:15:28,520
Speaker 1: way to act, might be UH, you know, a specific

230
00:15:29,200 --> 00:15:33,320
Speaker 1: um request, it might make whatever you may think, it

231
00:15:33,400 --> 00:15:38,480
Speaker 1: has to come to that conclusion, which means that we

232
00:15:38,560 --> 00:15:41,880
Speaker 1: don't know how robots are always going to react to

233
00:15:41,920 --> 00:15:46,320
Speaker 1: their environments. Environments can be very chaotic things with lots

234
00:15:46,320 --> 00:15:49,440
Speaker 1: of different variables. And while you might program a robot

235
00:15:49,480 --> 00:15:52,120
Speaker 1: so it will behave in a very predictable way for

236
00:15:52,240 --> 00:15:55,720
Speaker 1: certain situations, you're not gonna be able to predict every

237
00:15:55,760 --> 00:16:00,320
Speaker 1: possible situation situation in every possible environment. So because of that,

238
00:16:00,440 --> 00:16:05,200
Speaker 1: and because we're using machine learning in more applications, this

239
00:16:05,240 --> 00:16:08,600
Speaker 1: program was saying, or this this UH report was saying,

240
00:16:09,360 --> 00:16:12,640
Speaker 1: we need to keep that in mind. And while we

241
00:16:13,480 --> 00:16:17,240
Speaker 1: wouldn't necessarily call even the most advanced AI conscious or

242
00:16:17,240 --> 00:16:20,560
Speaker 1: self aware, it's not. It's not doesn't appear to be.

243
00:16:21,200 --> 00:16:23,840
Speaker 1: We are seeing more applications that allow machines to learn

244
00:16:23,840 --> 00:16:27,440
Speaker 1: from their environments and adapt their approaches to complete certain tasks,

245
00:16:27,560 --> 00:16:29,320
Speaker 1: and that is something we need to keep an eye

246
00:16:29,360 --> 00:16:31,920
Speaker 1: on now. Right now, you could argue that if a

247
00:16:32,040 --> 00:16:37,040
Speaker 1: robot malfunctions, you would say the company that made the

248
00:16:37,120 --> 00:16:41,320
Speaker 1: robot should be held liable or the programmer who programmed

249
00:16:41,360 --> 00:16:45,880
Speaker 1: in either the routine the robot was following or the

250
00:16:45,880 --> 00:16:51,280
Speaker 1: basic artificial intelligence that the robot would use as guideline. However,

251
00:16:51,440 --> 00:16:54,360
Speaker 1: as robots depend more and more upon machine learning to

252
00:16:54,440 --> 00:16:58,640
Speaker 1: interact with their environments, this gets really murky because again,

253
00:16:58,680 --> 00:17:02,680
Speaker 1: those environments are filled with predictable variables, and the robot

254
00:17:02,800 --> 00:17:05,280
Speaker 1: might quote unquote learned to do something in a way

255
00:17:05,320 --> 00:17:09,680
Speaker 1: that's harmful, or it's inefficient, or it's ineffective. This doesn't

256
00:17:09,720 --> 00:17:12,160
Speaker 1: necessarily mean it would cause harm to people. It might

257
00:17:12,280 --> 00:17:16,000
Speaker 1: just mess things up in like an industry environment and

258
00:17:16,119 --> 00:17:21,760
Speaker 1: cause financial harm. So the proposal argues that as machines

259
00:17:21,840 --> 00:17:27,640
Speaker 1: become more advanced, it makes less sense to blame the manufacturer.

260
00:17:27,720 --> 00:17:30,880
Speaker 1: It's less effective, and there might need to be new

261
00:17:30,960 --> 00:17:34,680
Speaker 1: rules and definitions for liability that would hold the machines

262
00:17:35,000 --> 00:17:41,399
Speaker 1: themselves responsible. So the proposal quote considers that a system

263
00:17:41,440 --> 00:17:45,399
Speaker 1: of registration of advanced robots should be introduced end quote.

264
00:17:45,640 --> 00:17:49,400
Speaker 1: That would involve creating the criteria to decide which robots

265
00:17:49,440 --> 00:17:52,720
Speaker 1: would be required for registration, and it would also call

266
00:17:52,760 --> 00:17:56,200
Speaker 1: for more funding from the European Union for research projects,

267
00:17:56,440 --> 00:18:01,080
Speaker 1: particularly those that involve the social and ethical challenges raised

268
00:18:01,119 --> 00:18:04,639
Speaker 1: by advances in robotics, and that the EU should create

269
00:18:04,720 --> 00:18:09,000
Speaker 1: a quote Legislative Instrument on legal questions related to the

270
00:18:09,000 --> 00:18:12,240
Speaker 1: development of robotics and AI end quote that would look

271
00:18:12,280 --> 00:18:16,560
Speaker 1: ahead ten to fifteen years, which is super hard to do.

272
00:18:17,200 --> 00:18:20,680
Speaker 1: It's really hard to predict what will happen in technology

273
00:18:20,720 --> 00:18:24,480
Speaker 1: in five years, let alone ten or fifteen. But they're

274
00:18:24,480 --> 00:18:27,840
Speaker 1: saying you should try and consider as many different possible

275
00:18:27,880 --> 00:18:33,320
Speaker 1: situations as you can. So why are they saying this, Well,

276
00:18:33,840 --> 00:18:38,320
Speaker 1: if I create a machine learning program and I have

277
00:18:38,400 --> 00:18:42,280
Speaker 1: a robot that follows it, and the robot is put

278
00:18:42,320 --> 00:18:45,200
Speaker 1: into an environment that's got a lot of these variables,

279
00:18:46,000 --> 00:18:48,000
Speaker 1: I may not know how my robots going to react

280
00:18:48,000 --> 00:18:51,760
Speaker 1: in every single situation. The more advanced the robot is,

281
00:18:52,480 --> 00:18:55,120
Speaker 1: the less certain I can be of exactly what it's

282
00:18:55,119 --> 00:19:01,120
Speaker 1: going to do based upon any given scenario. I might

283
00:19:01,480 --> 00:19:04,640
Speaker 1: feel like the guidelines I've created are enough to keep

284
00:19:04,680 --> 00:19:07,280
Speaker 1: the robot out of trouble. But the world is a

285
00:19:07,359 --> 00:19:10,800
Speaker 1: chaotic place, and we've talked about this with autonomous cars.

286
00:19:10,840 --> 00:19:12,600
Speaker 1: I'm actually going to talk about it more. I plan

287
00:19:12,680 --> 00:19:16,520
Speaker 1: on doing a whole suite of episodes about the history

288
00:19:17,320 --> 00:19:21,960
Speaker 1: an evolution and challenges of autonomous cars in the near future.

289
00:19:22,000 --> 00:19:27,600
Speaker 1: That's gonna be a multi part series. But with autonomous cars,

290
00:19:28,280 --> 00:19:34,480
Speaker 1: we know that there are situations that can end in accidents,

291
00:19:34,600 --> 00:19:39,240
Speaker 1: even fatal ones. We've seen examples of that, and that

292
00:19:39,760 --> 00:19:45,240
Speaker 1: this raises questions of what ultimately is what should we

293
00:19:45,280 --> 00:19:51,200
Speaker 1: hold accountable, who who is responsible for this? With driving scenarios,

294
00:19:51,640 --> 00:19:55,800
Speaker 1: we know we can predict maybe let's say nine of

295
00:19:55,840 --> 00:19:58,520
Speaker 1: the different scenarios you would encounter on a typical drive,

296
00:19:58,920 --> 00:20:00,680
Speaker 1: but that means there's ten percent and the stuff out

297
00:20:00,720 --> 00:20:04,200
Speaker 1: there that happens, it's just not normal. It's not something

298
00:20:04,200 --> 00:20:09,000
Speaker 1: you would typically encounter. And these outside scenarios pose a

299
00:20:09,040 --> 00:20:12,840
Speaker 1: problem because you cannot anticipate every single one in program

300
00:20:12,960 --> 00:20:16,920
Speaker 1: into your machine when this happens, make sure you do

301
00:20:17,000 --> 00:20:21,080
Speaker 1: this other thing. The machine will encounter scenarios in which

302
00:20:21,119 --> 00:20:24,320
Speaker 1: it has to make a decision for itself, and it's

303
00:20:24,359 --> 00:20:28,959
Speaker 1: at that point that we don't know where responsibility falls,

304
00:20:29,040 --> 00:20:34,080
Speaker 1: where the accountability falls because the programmer could not possibly

305
00:20:34,119 --> 00:20:37,400
Speaker 1: have anticipated this, So is it really fair to hold

306
00:20:37,440 --> 00:20:41,480
Speaker 1: them at fault. The manufacturer may have made it exactly

307
00:20:41,480 --> 00:20:43,359
Speaker 1: the way it was supposed to be made. There are

308
00:20:43,359 --> 00:20:46,360
Speaker 1: no faults in the manufacturing process, So are they at

309
00:20:46,359 --> 00:20:50,240
Speaker 1: fault the owner of the robot that put it in

310
00:20:50,640 --> 00:20:54,320
Speaker 1: whatever situation it found itself in the first place, maybe

311
00:20:54,359 --> 00:20:57,959
Speaker 1: they're at fault, or maybe the robot itself is at fault.

312
00:20:58,880 --> 00:21:02,679
Speaker 1: So this proposal goes on to call for a legal

313
00:21:02,720 --> 00:21:07,120
Speaker 1: solution that doesn't restrict the type or extent of damage

314
00:21:07,160 --> 00:21:10,080
Speaker 1: as a person can seek based solely on the fact

315
00:21:10,160 --> 00:21:13,120
Speaker 1: that the damage was caused by a non human agent. So,

316
00:21:13,359 --> 00:21:16,040
Speaker 1: in other words, a court should should not be allowed

317
00:21:16,040 --> 00:21:19,600
Speaker 1: to say, well, we can only award you this amount.

318
00:21:19,680 --> 00:21:22,960
Speaker 1: We can't give you any more than this because the

319
00:21:23,000 --> 00:21:25,760
Speaker 1: thing that caused harm to you was a robot, not

320
00:21:25,840 --> 00:21:27,920
Speaker 1: a person. If it were a person, we would be

321
00:21:27,960 --> 00:21:30,760
Speaker 1: able to award you more money. The proposal says, we

322
00:21:31,040 --> 00:21:33,280
Speaker 1: want to make sure that does not become the case.

323
00:21:33,800 --> 00:21:36,960
Speaker 1: And like I said, we're already seeing this in cases

324
00:21:36,960 --> 00:21:41,600
Speaker 1: with autonomous cars, where a company can say, well, you

325
00:21:41,640 --> 00:21:45,760
Speaker 1: know it was it was not the fault of the programmer,

326
00:21:45,880 --> 00:21:47,960
Speaker 1: was not the fault of the company. It was a

327
00:21:48,000 --> 00:21:51,000
Speaker 1: situation in which the car itself made that decision. The

328
00:21:51,080 --> 00:21:55,440
Speaker 1: car is at fault, not us. That is a problematic thing.

329
00:21:56,040 --> 00:21:59,639
Speaker 1: One of the things that the the report suggested was

330
00:22:00,119 --> 00:22:02,840
Speaker 1: that the producers of a robot are liable for damage

331
00:22:02,840 --> 00:22:05,439
Speaker 1: on a level that is proportionate with the amount of

332
00:22:05,480 --> 00:22:09,680
Speaker 1: instructions the producers gave to the robot. So the more

333
00:22:10,160 --> 00:22:13,920
Speaker 1: locked down the robot, the more responsible the producers are

334
00:22:14,000 --> 00:22:17,600
Speaker 1: for that robot's behavior. So if you have that stationary

335
00:22:17,880 --> 00:22:21,840
Speaker 1: robot that that is welding on an assembly line and

336
00:22:21,960 --> 00:22:25,879
Speaker 1: it malfunctions or it starts causing huge amounts of damage

337
00:22:25,880 --> 00:22:30,960
Speaker 1: and its behavior it's it's perhaps ruining several cars along

338
00:22:31,000 --> 00:22:33,800
Speaker 1: the assembly line as it's going through this process, you

339
00:22:33,840 --> 00:22:36,280
Speaker 1: would say, all right, the company that made this robot

340
00:22:36,400 --> 00:22:39,560
Speaker 1: is at fault. Something they did was wrong because the

341
00:22:39,640 --> 00:22:43,760
Speaker 1: robots only following the instructions that this company gave it.

342
00:22:43,760 --> 00:22:46,359
Speaker 1: It's not making any decisions on its own. But the

343
00:22:46,400 --> 00:22:50,439
Speaker 1: more the decision making process is on the robot, the

344
00:22:50,560 --> 00:22:54,639
Speaker 1: less you can hold the manufacturer accountable, according to this report,

345
00:22:54,640 --> 00:22:58,000
Speaker 1: which is a pretty radical idea, and the proposal also

346
00:22:58,040 --> 00:23:01,760
Speaker 1: states that the longer a robot has received quote unquote education,

347
00:23:02,160 --> 00:23:05,639
Speaker 1: the more liable the robots quote unquote teacher is for

348
00:23:05,680 --> 00:23:08,720
Speaker 1: any damage the robot causes. Now, that might be the

349
00:23:08,760 --> 00:23:10,680
Speaker 1: company that made the robot. It might be the person

350
00:23:10,760 --> 00:23:13,440
Speaker 1: that programmed the robot, it might be the person who

351
00:23:13,640 --> 00:23:16,920
Speaker 1: purchased the robot and then put it in an environment

352
00:23:16,960 --> 00:23:20,359
Speaker 1: where it was learning, but it would be whoever was

353
00:23:20,440 --> 00:23:25,359
Speaker 1: introducing these scenarios to the robot. So that's a really

354
00:23:26,040 --> 00:23:28,880
Speaker 1: interesting concept. And again, this is just a proposal. It's

355
00:23:28,880 --> 00:23:32,440
Speaker 1: not like this has been enacted into law. Uh. They

356
00:23:32,520 --> 00:23:36,960
Speaker 1: also created a suggestion for an obligatory insurance scheme. This

357
00:23:37,080 --> 00:23:39,800
Speaker 1: is not that crazy. I mean, if you drive a

358
00:23:39,840 --> 00:23:42,600
Speaker 1: car in the United States, you have to have insurance.

359
00:23:43,240 --> 00:23:45,639
Speaker 1: It's a requirement by law. So it'd be similar to that,

360
00:23:45,680 --> 00:23:48,920
Speaker 1: except in this case, the producers of the robots would

361
00:23:48,920 --> 00:23:52,120
Speaker 1: pay out the insurance for the robots that it was creating.

362
00:23:52,640 --> 00:23:55,920
Speaker 1: And the report also suggested that there should be or

363
00:23:55,960 --> 00:24:01,080
Speaker 1: there could be, a compensation fund for the robots, which

364
00:24:01,119 --> 00:24:03,880
Speaker 1: sounds crazy. Like one of the things you think about

365
00:24:03,960 --> 00:24:07,199
Speaker 1: about automation is that, well, it reduces the need to

366
00:24:07,280 --> 00:24:11,080
Speaker 1: have paid employees. You've got robots, why do you need

367
00:24:11,119 --> 00:24:15,040
Speaker 1: to pay them. Well, it's not to reward the robot.

368
00:24:15,080 --> 00:24:17,480
Speaker 1: It's not to make the robot feel like it did

369
00:24:17,480 --> 00:24:21,359
Speaker 1: a great job. Robots can't feel anything anyway. It's not

370
00:24:21,560 --> 00:24:26,520
Speaker 1: even to you know, to to have this system, you know,

371
00:24:26,680 --> 00:24:29,960
Speaker 1: perpetuate the system of paying for work. It's rather to

372
00:24:30,160 --> 00:24:34,119
Speaker 1: build a fund, a compensation fund that could cover the

373
00:24:34,160 --> 00:24:37,840
Speaker 1: cost of any damages or harm that the robot might create.

374
00:24:38,720 --> 00:24:42,280
Speaker 1: So you're not paying the robot and putting money into

375
00:24:42,440 --> 00:24:44,760
Speaker 1: its bank account that then goes to spend on motor

376
00:24:44,800 --> 00:24:48,600
Speaker 1: oil or something. You're accumulating money in the event that

377
00:24:48,640 --> 00:24:52,600
Speaker 1: the robot causes damage, and then you've got money dedicated

378
00:24:52,640 --> 00:24:55,320
Speaker 1: to that robot that you can use to pay out

379
00:24:55,840 --> 00:25:00,400
Speaker 1: and the event of damages being created by that robot. Uh,

380
00:25:00,640 --> 00:25:02,800
Speaker 1: so you're essentially paying a robot so that in case

381
00:25:02,840 --> 00:25:06,399
Speaker 1: it goes haywire and starts, I don't know, slapping people around,

382
00:25:06,720 --> 00:25:10,240
Speaker 1: you can actually cover all those damages. The report also

383
00:25:10,280 --> 00:25:14,600
Speaker 1: calls for specific legal status to apply to robots, thus

384
00:25:14,640 --> 00:25:19,679
Speaker 1: creating assas that's equivalent to like electronic persons that have

385
00:25:19,800 --> 00:25:23,080
Speaker 1: specific rights and obligations, including that of making good any

386
00:25:23,119 --> 00:25:26,440
Speaker 1: damage they may cause. And a lot of people say

387
00:25:26,520 --> 00:25:29,359
Speaker 1: that this is confusing, that why would you grant this

388
00:25:29,440 --> 00:25:33,159
Speaker 1: personhood to robots? And a frequent response to that, and

389
00:25:33,200 --> 00:25:35,280
Speaker 1: I think it's got some validity to it as well.

390
00:25:35,320 --> 00:25:38,440
Speaker 1: We already do it with businesses. We all are already

391
00:25:38,480 --> 00:25:43,640
Speaker 1: grant the concept of personhood to corporations. Corporations can behave

392
00:25:43,720 --> 00:25:48,119
Speaker 1: in legal matters as if they are people, human being people.

393
00:25:48,680 --> 00:25:51,359
Speaker 1: And if a corporation can do it a business a

394
00:25:51,560 --> 00:25:55,880
Speaker 1: collection of work, then why not a robot. Then there's

395
00:25:55,880 --> 00:25:58,399
Speaker 1: a section from the design of an Ethical Framework to

396
00:25:58,440 --> 00:26:01,840
Speaker 1: make certain advances that uh that are made in robotics

397
00:26:01,960 --> 00:26:08,320
Speaker 1: are made with consideration to how it impacts human safety, privacy, dignity,

398
00:26:08,840 --> 00:26:12,600
Speaker 1: your who owns information. This proposal goes on to say

399
00:26:12,640 --> 00:26:15,280
Speaker 1: that the risk of harm should be no greater than

400
00:26:15,480 --> 00:26:18,919
Speaker 1: encountered in ordinary life. So, in other words, a future

401
00:26:18,920 --> 00:26:21,359
Speaker 1: filled with robots should pose no more risk to a

402
00:26:21,440 --> 00:26:25,920
Speaker 1: person as that person would encounter today in ordinary circumstances. Now,

403
00:26:25,920 --> 00:26:28,679
Speaker 1: to do all of that, the proposals just creating a

404
00:26:28,760 --> 00:26:33,160
Speaker 1: new European Agency for Robotics and Artificial Intelligence and funding

405
00:26:33,200 --> 00:26:36,080
Speaker 1: that agency and staffing it with technical experts as well

406
00:26:36,119 --> 00:26:40,080
Speaker 1: as leaders in ethical and regulatory fields to start really

407
00:26:40,119 --> 00:26:43,680
Speaker 1: susten this stuff out. The other sections in the report

408
00:26:44,160 --> 00:26:48,280
Speaker 1: cover concepts like intellectual property, um, how do you protect

409
00:26:48,440 --> 00:26:52,200
Speaker 1: and encourage innovation? How do you call on new criteria

410
00:26:52,320 --> 00:26:55,440
Speaker 1: to apply to copyrightable works that are produced by computers

411
00:26:55,520 --> 00:26:58,800
Speaker 1: or robots. You may have listened to my episode, my

412
00:26:58,880 --> 00:27:04,480
Speaker 1: recent episode about the artificial intelligence that created a painting.

413
00:27:04,880 --> 00:27:09,440
Speaker 1: Who owns that painting? Does the machine own that painting?

414
00:27:09,520 --> 00:27:12,720
Speaker 1: We don't have rules for this, and that's what this

415
00:27:12,920 --> 00:27:16,760
Speaker 1: report is arguing is as we create more machines and

416
00:27:16,840 --> 00:27:21,440
Speaker 1: systems that can produce stuff that we would normally protect

417
00:27:22,280 --> 00:27:25,760
Speaker 1: with stuff like a copyright, what do we do? Who

418
00:27:25,840 --> 00:27:28,960
Speaker 1: owns that? Who should that go to? They also argue

419
00:27:28,960 --> 00:27:33,000
Speaker 1: for standardization, the idea of making sure that these robots

420
00:27:33,000 --> 00:27:37,440
Speaker 1: are all communicating with their various systems in a standardized way,

421
00:27:37,720 --> 00:27:39,480
Speaker 1: so that way, when you move from one part of

422
00:27:39,520 --> 00:27:45,120
Speaker 1: the EU to another, you don't have these conflicting communications standards,

423
00:27:45,800 --> 00:27:49,959
Speaker 1: data transfer standards. You want everything to work within the

424
00:27:50,040 --> 00:27:54,440
Speaker 1: same umbrella of standards, sort of avoiding the problem that

425
00:27:54,520 --> 00:27:57,000
Speaker 1: people often have, which is that you go to a

426
00:27:57,040 --> 00:28:02,800
Speaker 1: different country and everyone there has this audacity to talk

427
00:28:02,800 --> 00:28:05,560
Speaker 1: in a different language than you do, which means that

428
00:28:05,600 --> 00:28:08,520
Speaker 1: you have to speak really slow and loud if you're

429
00:28:08,560 --> 00:28:12,480
Speaker 1: an American. That's more of a joke about Americans being

430
00:28:12,640 --> 00:28:15,639
Speaker 1: rude and foreign nations. But you don't want that to

431
00:28:15,760 --> 00:28:18,080
Speaker 1: happen if you can prevent it, if you can create

432
00:28:18,080 --> 00:28:21,119
Speaker 1: the standards in the first place. They also call for

433
00:28:21,160 --> 00:28:25,119
Speaker 1: more standards to allow for the testing of driverless cars

434
00:28:25,880 --> 00:28:29,159
Speaker 1: and other autonomous vehicles rather than just this fragmented approach

435
00:28:29,200 --> 00:28:32,480
Speaker 1: that we're seeing, and also had a section on care

436
00:28:32,600 --> 00:28:36,919
Speaker 1: robots and medical robots general call to develop those robots

437
00:28:36,960 --> 00:28:40,680
Speaker 1: with a human impact in mind, and also the concept

438
00:28:40,720 --> 00:28:44,840
Speaker 1: of human repair and enhancement, which gets into this idea

439
00:28:44,920 --> 00:28:51,960
Speaker 1: of using robotics to either heal injury for people or

440
00:28:52,000 --> 00:28:56,120
Speaker 1: to maybe even replace people parts to make people cyborgs.

441
00:28:57,040 --> 00:29:00,160
Speaker 1: That's kind of a far off sort of thing that

442
00:29:00,200 --> 00:29:03,040
Speaker 1: we could think about. It may never happen, but the

443
00:29:03,120 --> 00:29:08,160
Speaker 1: report actually calls for a committee on robot ethics in

444
00:29:08,240 --> 00:29:11,280
Speaker 1: hospitals and other healthcare institutions to kind of develop ethical

445
00:29:11,320 --> 00:29:14,320
Speaker 1: guidelines for how that might be used. And then there's

446
00:29:14,360 --> 00:29:17,920
Speaker 1: a short section about drones, which is becoming more and

447
00:29:17,960 --> 00:29:21,320
Speaker 1: more important. I've got more to say about this, but

448
00:29:21,600 --> 00:29:31,840
Speaker 1: first let's take another quick break to thank our sponsor. Now,

449
00:29:31,840 --> 00:29:36,640
Speaker 1: according to one forecast that the reports sites, the European

450
00:29:36,760 --> 00:29:40,440
Speaker 1: Union could face a shortage of nearly a million information

451
00:29:40,560 --> 00:29:44,440
Speaker 1: and communications technology professionals in the near future. And on

452
00:29:44,520 --> 00:29:47,920
Speaker 1: top of that, it says that of all jobs will

453
00:29:48,000 --> 00:29:52,120
Speaker 1: require some level of digital skills moving forward, so it

454
00:29:52,200 --> 00:29:56,080
Speaker 1: calls for a revision of a digital competence framework to

455
00:29:56,240 --> 00:29:59,720
Speaker 1: correct for that, and it also calls for designing programs

456
00:29:59,720 --> 00:30:03,480
Speaker 1: doing courage people who typically don't go into these fields

457
00:30:03,840 --> 00:30:07,920
Speaker 1: to pursue them, specifically young women, to get young women

458
00:30:07,920 --> 00:30:11,760
Speaker 1: into the field of robotics, and that the European Union

459
00:30:11,760 --> 00:30:14,880
Speaker 1: and member states should quote launch initiatives in order to

460
00:30:14,960 --> 00:30:18,360
Speaker 1: support women in i c T that's the information and

461
00:30:18,360 --> 00:30:22,360
Speaker 1: communications technology and to boost their E skills end quote.

462
00:30:23,200 --> 00:30:26,240
Speaker 1: They also call for a system to monitor job trends

463
00:30:26,320 --> 00:30:30,120
Speaker 1: to see where jobs are disappearing due to automation, where

464
00:30:30,200 --> 00:30:34,800
Speaker 1: jobs are being created because of robotics, and to stay

465
00:30:34,840 --> 00:30:37,360
Speaker 1: ahead of that to say, well, we see where things

466
00:30:37,360 --> 00:30:40,840
Speaker 1: are going, so we know what to stress to people

467
00:30:40,880 --> 00:30:44,440
Speaker 1: who are in school, like what areas of opportunity are there,

468
00:30:44,480 --> 00:30:48,000
Speaker 1: so that people will go into those areas one to

469
00:30:48,080 --> 00:30:52,000
Speaker 1: meet the demand that's going to be created, and to

470
00:30:52,000 --> 00:30:56,120
Speaker 1: to avoid going into fields that would largely be obsolete,

471
00:30:56,360 --> 00:30:59,240
Speaker 1: meaning that they would have difficulty finding work after they

472
00:30:59,480 --> 00:31:03,920
Speaker 1: came out of the education system, and maybe that the

473
00:31:04,000 --> 00:31:08,400
Speaker 1: EU should introduce corporate reporting requirements on the extent and

474
00:31:08,440 --> 00:31:11,680
Speaker 1: proportion of the contribution of robotics and AI to the

475
00:31:11,680 --> 00:31:15,840
Speaker 1: economic results of a company for the purpose of taxation

476
00:31:16,200 --> 00:31:21,640
Speaker 1: and social security contributions. So not only does this report

477
00:31:21,680 --> 00:31:27,200
Speaker 1: suggest that employers could quote unquote pay robots, although again

478
00:31:27,240 --> 00:31:29,560
Speaker 1: that would be to put money aside in the event

479
00:31:29,680 --> 00:31:32,760
Speaker 1: that something were to go catastrophically wrong with that robot,

480
00:31:33,480 --> 00:31:36,920
Speaker 1: but that the robots would also have to pay into

481
00:31:36,920 --> 00:31:40,760
Speaker 1: the social security system, or rather the company's employing those

482
00:31:40,880 --> 00:31:45,240
Speaker 1: robots would need to And the report is saying that

483
00:31:45,440 --> 00:31:49,560
Speaker 1: if it is a fact that overall we're going to

484
00:31:49,600 --> 00:31:54,200
Speaker 1: see a decrease in employment because of automation, that would

485
00:31:54,240 --> 00:31:57,440
Speaker 1: be really bad. It would have a domino effect ripple

486
00:31:57,480 --> 00:32:00,560
Speaker 1: effect on stuff like Social Security because it depends on

487
00:32:00,600 --> 00:32:05,440
Speaker 1: employment taxes, and without those taxes, systems like Social Security

488
00:32:05,520 --> 00:32:09,320
Speaker 1: would lose funding and people would be and put in

489
00:32:09,440 --> 00:32:13,360
Speaker 1: hardship because of that. So there would need to be

490
00:32:13,400 --> 00:32:16,840
Speaker 1: some sort of tax on robots to help compensate for this,

491
00:32:17,240 --> 00:32:19,800
Speaker 1: to to level out the fact that there would not

492
00:32:19,800 --> 00:32:25,200
Speaker 1: be these employment taxes that employees would typically be UH

493
00:32:25,440 --> 00:32:28,560
Speaker 1: paying for out of their salaries. So the proposal also

494
00:32:28,600 --> 00:32:32,120
Speaker 1: suggests that all members of the EU consider a general

495
00:32:32,200 --> 00:32:37,560
Speaker 1: basic income as a possibility in case this wave of

496
00:32:37,600 --> 00:32:41,880
Speaker 1: automation does take large effect, because otherwise you could have

497
00:32:42,160 --> 00:32:45,440
Speaker 1: a population that's largely out of work and has no

498
00:32:45,520 --> 00:32:48,120
Speaker 1: way of making a living. A general basic income, like

499
00:32:48,160 --> 00:32:52,000
Speaker 1: a guaranteed basic income, would be an amount of money

500
00:32:52,080 --> 00:32:54,240
Speaker 1: that the government would pay out to each and every

501
00:32:54,280 --> 00:32:58,480
Speaker 1: citizen that would be used to cover basic needs. It

502
00:32:58,520 --> 00:33:00,840
Speaker 1: would not prevent people from going out and getting a

503
00:33:00,960 --> 00:33:04,120
Speaker 1: job and earning more income that they could do if

504
00:33:04,200 --> 00:33:07,800
Speaker 1: they wanted to live above the basic line that had

505
00:33:07,840 --> 00:33:11,160
Speaker 1: been set by whatever the basic income amount was. But

506
00:33:11,200 --> 00:33:13,240
Speaker 1: the idea would be that the basic income would would

507
00:33:13,240 --> 00:33:17,760
Speaker 1: cover your your most important needs like a place to sleep, food,

508
00:33:18,200 --> 00:33:20,760
Speaker 1: that kind of stuff. The proposal also lays out some

509
00:33:20,760 --> 00:33:24,959
Speaker 1: guidelines for licenses. One set that would be meant for

510
00:33:25,480 --> 00:33:28,760
Speaker 1: designers of these robots, one set that would meant be

511
00:33:28,840 --> 00:33:32,360
Speaker 1: meant for the users of the robots. UH. They do

512
00:33:32,520 --> 00:33:37,000
Speaker 1: have a few interesting examples in the report. There's actually

513
00:33:37,040 --> 00:33:38,920
Speaker 1: quite a few examples, but some of the fun ones

514
00:33:38,920 --> 00:33:42,920
Speaker 1: like under designers there's UH integrate opt out mechanisms like

515
00:33:43,120 --> 00:33:47,240
Speaker 1: kill switches consistent with design objectives. Make sure your robot

516
00:33:47,320 --> 00:33:50,600
Speaker 1: is going to operate in a legal way, so you know,

517
00:33:50,880 --> 00:33:54,120
Speaker 1: don't go making a gangster bot that would be dumb.

518
00:33:54,960 --> 00:33:57,800
Speaker 1: Be transparent in the way that the robot is programmed,

519
00:33:57,840 --> 00:34:00,960
Speaker 1: as well as the predictability of its robotic behavior so

520
00:34:01,000 --> 00:34:04,280
Speaker 1: people know what to expect when they use it. Develop

521
00:34:04,560 --> 00:34:07,520
Speaker 1: tracing tools during the development stages so that when a

522
00:34:07,600 --> 00:34:10,200
Speaker 1: robot behaves a particular way, it can be traced back

523
00:34:10,239 --> 00:34:13,520
Speaker 1: to the design of the robot, and doing that would

524
00:34:13,520 --> 00:34:16,920
Speaker 1: help other designers either incorporate good design elements into their

525
00:34:16,920 --> 00:34:22,040
Speaker 1: approach or avoid designs that lead to you know, the

526
00:34:22,440 --> 00:34:26,120
Speaker 1: crazy Viking robots that go on rampages in Northern Europe.

527
00:34:26,520 --> 00:34:29,280
Speaker 1: Make sure that the robots are identifiable as robots as well.

528
00:34:29,640 --> 00:34:31,719
Speaker 1: That was a big one they said that should be

529
00:34:31,800 --> 00:34:34,560
Speaker 1: part of it. People should know when they're looking at

530
00:34:34,560 --> 00:34:37,640
Speaker 1: a robot that it is in fact a robot. Then

531
00:34:37,719 --> 00:34:40,360
Speaker 1: for users, some of the things that were in the

532
00:34:40,400 --> 00:34:45,160
Speaker 1: licenses included respect human physical and emotional frailty, so you

533
00:34:45,200 --> 00:34:51,360
Speaker 1: should not have you know, robots employed as as strike breakers.

534
00:34:52,040 --> 00:34:55,640
Speaker 1: You should respect other people's privacy, which includes turning off

535
00:34:55,680 --> 00:34:58,719
Speaker 1: a robots video recording equipment if the situation warrants it,

536
00:34:59,440 --> 00:35:05,440
Speaker 1: and not to weaponize robots, which seems pretty you know, straightforward. Now,

537
00:35:05,560 --> 00:35:08,200
Speaker 1: this was just a proposal, and it's still something that

538
00:35:08,320 --> 00:35:13,080
Speaker 1: is being debated. In the European Union. There has undergone

539
00:35:13,640 --> 00:35:16,799
Speaker 1: rewrites and tweaks since it was first proposed a couple

540
00:35:16,800 --> 00:35:20,280
Speaker 1: of years ago, and in the EU there are still

541
00:35:20,600 --> 00:35:24,680
Speaker 1: discussions that are happening happening regularly about you know, what,

542
00:35:24,760 --> 00:35:29,160
Speaker 1: if anything, should the EU do about the development of

543
00:35:29,200 --> 00:35:33,400
Speaker 1: machine learning, artificial intelligence and automation, and what are the

544
00:35:33,440 --> 00:35:36,120
Speaker 1: best courses of action. I think the report was a

545
00:35:36,120 --> 00:35:39,920
Speaker 1: particularly interesting proposal. I think it had a lot of

546
00:35:40,040 --> 00:35:43,080
Speaker 1: very ambitious parts to it. I don't know that all

547
00:35:43,120 --> 00:35:46,440
Speaker 1: of them were even remotely realistic. Some of them certainly,

548
00:35:46,560 --> 00:35:49,239
Speaker 1: but I don't know about all of them. But I

549
00:35:49,680 --> 00:35:51,759
Speaker 1: thought that the most important thing was that it would

550
00:35:51,800 --> 00:35:55,280
Speaker 1: get people talking. The problem is we're still talking because

551
00:35:55,480 --> 00:35:59,560
Speaker 1: it's a complicated thing to think about and it requires

552
00:35:59,600 --> 00:36:05,120
Speaker 1: a lot of subtle decision making that it's just not

553
00:36:05,200 --> 00:36:11,799
Speaker 1: easy to do. It requires very careful consideration, and things

554
00:36:11,840 --> 00:36:15,000
Speaker 1: are changing so quickly that it can be difficult to

555
00:36:15,640 --> 00:36:18,480
Speaker 1: even get a handle on what's happening right now, let

556
00:36:18,520 --> 00:36:23,120
Speaker 1: alone figure out what might happen in ten years. It

557
00:36:23,200 --> 00:36:25,840
Speaker 1: does not change the fact that we need to consider

558
00:36:25,920 --> 00:36:29,480
Speaker 1: these things and to come up with some solutions. They

559
00:36:29,560 --> 00:36:32,839
Speaker 1: might not require granting personhood to robots. That might be

560
00:36:33,080 --> 00:36:36,839
Speaker 1: a step too far, but it would be really good

561
00:36:36,880 --> 00:36:40,680
Speaker 1: to get a stronger handle on what is coming down

562
00:36:40,840 --> 00:36:44,719
Speaker 1: the pipeline so that we're not blindsided by it and

563
00:36:44,840 --> 00:36:47,800
Speaker 1: we can limit the negative impact, if any, that it

564
00:36:47,840 --> 00:36:50,799
Speaker 1: would have on people in general. And I just think

565
00:36:50,800 --> 00:36:53,719
Speaker 1: it's fascinating this idea that we are really having these

566
00:36:53,719 --> 00:36:59,040
Speaker 1: conversations about how do we go forward as automation and

567
00:36:59,160 --> 00:37:02,680
Speaker 1: AI and machine learning become more and more part of

568
00:37:02,760 --> 00:37:05,880
Speaker 1: our lives, even if it's not in ways we directly

569
00:37:05,960 --> 00:37:08,719
Speaker 1: observe on a day to day basis. That wraps up

570
00:37:08,760 --> 00:37:11,120
Speaker 1: this episode. Like I said, pretty soon, I'm gonna do

571
00:37:11,239 --> 00:37:16,120
Speaker 1: a suite on autonomous cars and talk about their development

572
00:37:16,200 --> 00:37:19,919
Speaker 1: and the technology behind them, the ethical issues there. We'll

573
00:37:19,920 --> 00:37:22,640
Speaker 1: get into the trolley problem. It's one of my favorite

574
00:37:22,680 --> 00:37:26,680
Speaker 1: logical problems or ethical problems to discuss. I talked about

575
00:37:26,719 --> 00:37:29,320
Speaker 1: a little bit recently. We'll get into more detail about

576
00:37:29,400 --> 00:37:32,279
Speaker 1: that because people are having to talk about it now

577
00:37:32,760 --> 00:37:36,560
Speaker 1: and it's it's potentially scary stuff but also really fascinating.

578
00:37:36,840 --> 00:37:39,600
Speaker 1: If you guys have any suggestions for future episodes of

579
00:37:39,640 --> 00:37:42,600
Speaker 1: tech Stuff, maybe there's a technology you want me to cover,

580
00:37:42,760 --> 00:37:45,200
Speaker 1: a company, maybe there's someone you would like me to

581
00:37:45,200 --> 00:37:48,200
Speaker 1: talk about, or to go on over to tech Stuff

582
00:37:48,239 --> 00:37:51,000
Speaker 1: podcast dot com. That's the website. You'll find all the

583
00:37:51,000 --> 00:37:53,680
Speaker 1: ways to contact me there. I look forward to hearing

584
00:37:53,680 --> 00:37:55,560
Speaker 1: from you. Also, make sure you head over to t

585
00:37:55,800 --> 00:37:58,960
Speaker 1: public dot com slash tech Stuff. That's our merchandise store.

586
00:37:58,960 --> 00:38:01,239
Speaker 1: You're gonna find all sorts tech stuff merch over there.

587
00:38:01,600 --> 00:38:04,560
Speaker 1: There's cool things to check out. We're gonna be adding

588
00:38:04,600 --> 00:38:08,399
Speaker 1: some more designs in the very near future, and every

589
00:38:08,440 --> 00:38:10,960
Speaker 1: purchase you make goes to help the show, so we

590
00:38:11,080 --> 00:38:15,879
Speaker 1: greatly appreciate you supporting us. It's fantastic and I will

591
00:38:15,920 --> 00:38:24,080
Speaker 1: talk to you again really soon for more on this

592
00:38:24,280 --> 00:38:26,800
Speaker 1: and thousands of other topics because it how stuff works.

593
00:38:26,800 --> 00:38:36,960
Speaker 1: Dot com